Do Method and Lack of Context Explain Low Stability of Brand Associations? No!

Abstract

Scholars report low stability of brand-attribute associations in free choice affirmative binary rating tasks (“yes”-option, ticked or not, indicating whether a brand has a specific attribute or not). A multi-brand meta-analysis by Rungie, Laurent, Rily, Morrison and Roy (2005) reports stability levels of 49% over a period of three months; another by Dolnicar and Rossiter (2008) an average of 53% for 20 brands with a one-week interval between two measurements. Dolnicar and Rossiter (2008) provide a set of recommendations to attain higher levels of stability, for example shorter questionnaires, or sampling only category users and brand aware respondents.